CN115604765A - Computing offload optimization method and device, electronic equipment and storage medium - Google Patents

Abstract

The present application provides a computing offloading optimization method and device, electronic equipment and storage media, wherein the method includes: establishing a blockchain-based cloud native network system; The task arrival rate determines the reputation of each pair of alliance nodes; generates a blockchain consensus mechanism based on the reputation; obtains the first time delay and the first energy consumption required by the alliance chain system in the calculation offloading process, and obtains the blockchain The second delay required by the consensus mechanism in the consensus process; establish the delay minimization target model based on the first delay, the first energy consumption and the second delay, and obtain the calculation unloading optimization scheme based on the delay minimization target model . Through this application, the problem in the related art that cannot simultaneously meet the computing real-time and data security requirements of the Internet of Things device is solved.

Description

Calculation offload optimization method and device, electronic device and storage medium

technical field

The present invention relates to the field of computer technology, in particular to a computing offloading optimization method and device, electronic equipment and a storage medium.

Background technique

In recent years, the explosive growth of IoT devices has prompted the continuous emergence of computing-intensive applications, such as: intelligent driving, virtual/augmented reality, and online games. These applications have the characteristics of large amount of calculation and high real-time performance. However, IoT devices have limited computing and storage resources, which cannot meet the ultra-low latency service requirements of computing-intensive applications. Using cloud-native network architecture, with the cloud as the center of the network, all services are directly connected to the cloud, making the network evolve from the traditional end-to-end transmission connection to the efficient collaboration between the cloud and the end. The edge cloud provides computing and storage resources, and IoT devices can offload computing tasks to edge cloud servers for execution, thereby effectively reducing the computing delay of tasks and the energy consumption of IoT devices. However, because edge devices are usually deployed at the edge of networks such as wireless base stations, they are distributed and heterogeneous, and they are in an open and insecure environment, which leads to data security issues in the computing offload of IoT devices.

Related technologies introduce blockchain technology to ensure the data security of IoT devices during task offloading. As a multi-party co-construction, sharing and co-management technology, blockchain can ensure the safety and reliability of data during computing offloading through a consensus mechanism. However, the traditional blockchain consensus mechanism has problems such as low transaction throughput, large resource consumption, and prolonged transaction authentication time, which cannot guarantee data security when IoT devices with massive resource constraints frequently perform calculation offloading. Moreover, most of the optimization goals of existing computing offloading methods only consider reducing the delay caused by computing offloading tasks, without considering the transaction authentication delay of the blockchain system.

Therefore, the prior art has the problem of being unable to simultaneously meet the computing real-time and data security requirements of IoT devices.

Contents of the invention

The present application provides a calculation offloading optimization method and device, electronic equipment and storage media, so as to at least solve the problems in the related art that cannot satisfy both real-time calculation and data security of IoT devices.

According to an aspect of an embodiment of the present application, a calculation offloading optimization method is provided, the method comprising:

Establish a blockchain-based cloud-native network system, wherein the cloud-native network system includes an alliance chain system for information processing in the calculation offloading process, and the alliance chain system includes multiple pairs of alliance nodes, and each pair of alliance nodes Including a target IoT device and a target edge server, the target IoT device is an object for sending task offloading information, and the target edge server is an object for calculating the task offloading information and performing block chain consensus;

determining the credibility of each pair of federated nodes according to the edge computing results of the target edge server and the task arrival rate of the target IoT device;

Generate a blockchain consensus mechanism based on the reputation;

Obtain the first time delay and the first energy consumption required by the consortium chain system during the calculation offloading process, and obtain the second time delay required by the blockchain consensus mechanism during the consensus process;

A time delay minimization target model is established according to the first time delay, the first energy consumption, and the second time delay, and a computing offload optimization solution is obtained based on the time delay minimization target model.

According to another aspect of the embodiments of the present application, there is also provided a computing offloading optimization device, which includes:

Establishing a module for establishing a blockchain-based cloud-native network system, wherein the cloud-native network system includes an alliance chain system for information processing in the calculation offloading process, and the alliance chain system includes multiple pairs of alliance nodes, each pair The alliance node includes a target IoT device and a target edge server, the target IoT device is an object for sending task offload information, and the target edge server is an object for calculating the task offload information and performing block chain consensus;

A determination module, configured to determine the credibility of each pair of alliance nodes according to the edge computing results of the target edge server and the task arrival rate of the target IoT device;

A generating module, configured to generate a block chain consensus mechanism according to the credibility;

The obtaining module is used to obtain the first time delay and the first energy consumption required by the consortium chain system during the calculation offloading process, and obtain the second time delay required by the blockchain consensus mechanism during the consensus process ;

The obtaining module is configured to establish a delay minimization target model according to the first time delay, the first energy consumption, and the second time delay, and obtain a calculation unloading optimization scheme based on the time delay minimization target model.

Optionally, build modules include:

A creation unit is configured to use the target edge server to create an initial unit, wherein the initial unit is used to store the edge calculation result and the reputation of the alliance node;

A selection unit, configured to use the target edge server to select a first preset number of edge blocks through a first preset method, and store hash values of the first preset number of edge blocks into the initial unit , get the first unit;

a calculation unit, configured to use the target edge server to store a random number into the first unit, calculate a hash value of the first unit at this time, and store the hash value into the first unit, get the second unit;

a broadcast unit, configured to use the target edge server to broadcast the second unit to other edge servers;

A verification unit, configured to use the other edge server to verify whether the second unit is legal, and if legal, the second unit becomes a new edge block;

The judging unit is configured to use the edge blocks generated by the other edge servers to verify the new edge blocks, and use the other edge servers to judge whether the number of verification times reaches the authentication threshold, and if it reaches the authentication threshold, then The data of the second unit is successfully consensused by the consortium chain system.

Optionally, the acquisition module includes:

The first acquiring unit is configured to acquire the transmission delay during transmission of the task offloading information and the first energy consumption, the queuing delay of the task offloading information at the edge server and the edge server executing the Execution latency required for task offload information;

a first obtaining unit, configured to obtain the first time delay according to the transmission time delay, the queuing time delay and the execution time delay;

A second acquiring unit, configured to acquire the second time delay of the task offloading information in the consensus process.

Optionally, the resulting modules include:

a generating unit, configured to generate constraint conditions for the first time delay, the first energy consumption, and the second time delay;

A first establishing unit, configured to establish the delay minimization target model based on the constraint conditions.

a conversion unit, configured to convert the time-delay minimization target model into a second preset number of local models by a second preset method;

The second obtaining unit is configured to optimize the local models in parallel by training the parameters of the second preset number of local models to obtain corresponding local variables;

The third obtaining unit is configured to combine the local variables and optimize the global variables of the time delay minimization objective model to obtain the calculation unloading optimization scheme.

Optionally, the determination module includes:

The fourth obtaining unit is configured to use the target edge server to obtain the comprehensive reputation of the target edge server according to the reputation of a third preset number of the target IoT devices to the target edge server;

The fifth obtaining unit is configured to use the target edge server to obtain the reputation of the target IoT device according to the task arrival rate of the target IoT device;

A determining unit, configured to determine the reputation of each pair of alliance nodes according to the comprehensive reputation of the target edge server and the reputation of the target IoT device.

Optionally, building modules includes:

The second establishment unit is used to establish a blockchain-powered cloud-native network system, wherein the cloud-native network system includes a second preset number of IoT devices and a fourth preset number of edge servers;

The third establishing unit is configured to establish the alliance chain system in the calculation offloading process based on the cloud native network system.

According to yet another aspect of the embodiments of the present application, there is also provided an electronic device, including a processor, a communication interface, a memory, and a communication bus, wherein, the processor, the communication interface, and the memory complete mutual communication through the communication bus; wherein, The memory is used to store a computer program; the processor is used to execute the method steps in any one of the above embodiments by running the computer program stored in the memory.

According to still another aspect of the embodiments of the present application, there is also provided a computer-readable storage medium, in which a computer program is stored, wherein the computer program is set to execute the method in any of the above-mentioned embodiments when running step.

In the embodiment of this application, by establishing a blockchain-based cloud-native network system, the cloud-native network system includes an alliance chain system for information processing in the process of computing offload, and the alliance chain system includes multiple pairs of alliance nodes. The node includes the target IoT device and the target edge server. The target IoT device is the object that sends task offloading information, and the target edge server is the object that calculates task offloading information and performs blockchain consensus; according to the edge computing results of the target edge server and The task arrival rate of the target IoT device determines the reputation of each pair of alliance nodes; generates a blockchain consensus mechanism based on the reputation; obtains the first time delay and the first energy consumption required by the alliance chain system in the calculation offloading process 1. Obtain the second delay required by the blockchain consensus mechanism in the consensus process; establish a delay minimization target model based on the first delay, first energy consumption, and second delay, and based on the delay minimization target model Obtain the calculation offloading optimization scheme. Since the embodiment of this application, on the one hand, establishes a blockchain-based cloud-native network system and generates a consensus mechanism based on the credibility of alliance nodes, it reduces the transaction authentication delay and improves the security of the blockchain consensus. , optimize the calculation offloading task and the blockchain consensus task at the same time, realize the minimum system delay of the cloud-native network empowered by the blockchain, and solve the problems in related technologies that cannot satisfy the computing real-time and data requirements of IoT devices at the same time The question of security requirements.

Description of drawings

The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and together with the description serve to explain the principles of the invention.

In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings that need to be used in the description of the embodiments or the prior art. Obviously, for those of ordinary skill in the art, In other words, other drawings can also be obtained from these drawings without paying creative labor.

FIG. 1 is a schematic flowchart of an optional computing offloading optimization method according to an embodiment of the present application;

FIG. 2 is a model diagram of an optional blockchain-powered cloud-native network system according to an embodiment of the present application;

Fig. 3 is a structural block diagram of an optional computing offload optimization device according to an embodiment of the present application;

Fig. 4 is a structural block diagram of an optional electronic device according to an embodiment of the present application.

detailed description

In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiment of the application will be clearly and completely described below in conjunction with the accompanying drawings in the embodiment of the application. Obviously, the described embodiment is only It is an embodiment of a part of the application, but not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by persons of ordinary skill in the art without creative efforts shall fall within the scope of protection of this application.

It should be noted that the terms "first" and "second" in the description and claims of the present application and the above drawings are used to distinguish similar objects, but not necessarily used to describe a specific sequence or sequence. It is to be understood that the data so used are interchangeable under appropriate circumstances such that the embodiments of the application described herein can be practiced in sequences other than those illustrated or described herein. Furthermore, the terms "comprising" and "having", as well as any variations thereof, are intended to cover a non-exclusive inclusion, for example, a process, method, system, product or device comprising a sequence of steps or elements is not necessarily limited to the expressly listed Those steps or elements may instead include other steps or elements not explicitly listed or inherent to the process, method, product or apparatus.

In cloud-native networks, IoT devices with limited resources can reduce energy consumption and computing delay by offloading computing tasks to edge servers. Blockchain technology can ensure data security in the above process through a consensus mechanism. However, the traditional blockchain consensus mechanism has problems such as low transaction throughput, large resource consumption, and prolonged transaction authentication, which cannot guarantee data security when IoT devices with massive resource constraints frequently perform computing offloading tasks. In addition, the execution of the blockchain consensus mechanism and computing offloading tasks by the edge server will generate delays. If these two delays are optimized independently, it will not be able to meet the real-time calculation and data security service quality requirements of IoT devices at the same time. In order to solve the above problem, according to an aspect of the embodiment of the present application, a calculation offloading optimization method is provided, as shown in FIG. 1 , the process of the method may include the following steps:

Step S101, establish a blockchain-based cloud-native network system, wherein the cloud-native network system includes a consortium chain system for information processing in the process of computing offloading, and the consortium chain system includes multiple pairs of consortium nodes, each pair of consortium nodes contains target objects The networked device and the target edge server, the target IoT device is the object that sends task offloading information, and the target edge server is the object that calculates task offloading information and performs blockchain consensus.

Optionally, a blockchain-powered cloud-native network system is established. As shown in Figure 2, there are multiple IoT devices and multiple edge servers in the cloud-native network. In this cloud-native network system, an alliance chain system jointly maintained by two types of interest groups, IoT devices and edge servers, is built. In the consortium chain system, multiple consortium nodes are generated by multiple IoT devices and multiple edge servers, and an IoT device and an edge server selected by it can also be selected as a pair of consortium nodes. Among them, IoT devices, as blockchain users, send computing offloading task information. In addition to performing computing offloading tasks, edge servers, as blockchain consensus nodes, package edge computing results and other related information into blocks, and execute consensus algorithms to maintain The blockchain ledger ensures the service quality requirements of low energy consumption, low latency and high reliability of IoT devices.

Step S102, determine the credibility of each pair of alliance nodes according to the edge computing results of the target edge server and the task arrival rate of the target IoT device.

Optionally, for a blockchain system based on a directed acyclic graph (Directed Acyclic Graph, DAG), when the transaction arrival rate is low or unstable, the performance of the blockchain system will be reduced. The higher the task arrival rate, the more computing offloading tasks the edge server needs to perform, the more user data information needs to be protected, and the higher the transaction arrival rate of the blockchain. According to the edge computing results of the edge server, the IoT device evaluates the satisfaction of the edge server. The higher the satisfaction of the IoT device, the better the computing task of the edge server is completed, the higher the reliability of the edge server is, and the impact of the IoT device on the edge server is higher. The higher the credibility score. The longer the online time of the IoT device is, the higher the activity is, and the higher the credit score of the edge server is for the IoT device.

Step S103, generating a blockchain consensus mechanism according to the credibility.

Optionally, the DAG-based blockchain adopts an asynchronous communication mechanism, allowing the blockchain to perform bifurcation verification, requiring that after a new block enters the network, two existing blocks in the network must be randomly selected for verification. For an alliance node with a high reputation, the higher the probability of its published transaction being selected for verification, the faster the transaction verification speed, and the shorter the transaction verification delay. For alliance nodes with low credibility, the probability of their issued transactions being verified is low, the transaction verification speed is slow, and the transaction verification time is prolonged.

Step S104, obtain the first time delay required by the consortium chain system in the calculation offloading process and the first energy consumption consumed, and obtain the second time delay required by the blockchain consensus mechanism in the consensus process.

整个计算过程中的系统时延包括第一时延即任务卸载时延

和第二时延即区块链交易认证时延

任务卸载时延包括任务传输时延

排队时延

和边缘服务器的计算时延

由于边缘计算结果的数据大小与计算任务的数据大小相比非常小，因而不考虑计算结果传输的时延。另外，由于物联网设备的能量有限，边缘服务器有持续供电的电源，因而只考虑物联网设备的能耗，不考虑边缘服务器产生的能耗，因此第一能耗等于传输阶段的能耗

Alternatively, assume that the average data size and maximum allowable delay of each task d _i are D _i and

The system delay in the whole calculation process includes the first delay, which is the task offload delay

and the second delay is the blockchain transaction authentication delay

Task offload delay includes task transmission delay

queuing delay

and the computing delay of the edge server

Since the data size of edge computing results is very small compared with the data size of computing tasks, the delay in the transmission of computing results is not considered. In addition, due to the limited energy of IoT devices, edge servers have continuous power supply, so only the energy consumption of IoT devices is considered, and the energy consumption generated by edge servers is not considered, so the first energy consumption is equal to the energy consumption in the transmission stage

In step S105, a delay minimization target model is established according to the first time delay, the first energy consumption, and the second time delay, and an optimization solution for computing unloading is obtained based on the time delay minimization target model.

越低，共识安全性越高，但此时任务卸载时延

较高，不能满足物联网设备时延的服务质量要求。为了同时满足物联网设备计算实时性和数据安全性的服务质量需求，本申请实施例对计算卸载任务和区块共识任务同时进行优化，以最小化

和

的加权和为优化目标，建立一个数学优化模型作为时延最小化目标模型，并根据该数学优化模型得到计算卸载优化方案。Optionally, according to the characteristics of the DAG blockchain consensus mechanism, the higher the task arrival rate λ _i , the longer the blockchain transaction authentication delay

The lower the value, the higher the security of the consensus, but at this time the task unloading delay

It is relatively high and cannot meet the quality of service requirements for the delay of IoT devices. In order to meet the real-time calculation and data security quality of service requirements of IoT devices, the embodiment of this application optimizes the calculation offloading task and the block consensus task at the same time to minimize

and

The weighted sum of is the optimization objective, and a mathematical optimization model is established as the delay minimization objective model, and the calculation unloading optimization scheme is obtained according to the mathematical optimization model.

In the embodiment of this application, by establishing a blockchain-based cloud-native network system, the cloud-native network system includes an alliance chain system for information processing in the process of computing offload, and the alliance chain system includes multiple pairs of alliance nodes. The node includes the target IoT device and the target edge server. The target IoT device is the object that sends task offloading information, and the target edge server is the object that calculates task offloading information and performs blockchain consensus; according to the edge computing results of the target edge server and The task arrival rate of the target IoT device determines the reputation of each pair of alliance nodes; generates a blockchain consensus mechanism based on the reputation; obtains the first time delay and the first energy consumption required by the alliance chain system in the calculation offloading process 1. Obtain the second delay required by the blockchain consensus mechanism in the consensus process; establish a delay minimization target model based on the first delay, first energy consumption, and second delay, and based on the delay minimization target model Obtain the calculation offloading optimization scheme. Since the embodiment of this application, on the one hand, establishes a blockchain-based cloud-native network system and generates a consensus mechanism based on the credibility of alliance nodes, it reduces the transaction authentication delay and improves the security of the blockchain consensus. , optimize the calculation offloading task and the blockchain consensus task at the same time, realize the minimum system delay of the cloud-native network empowered by the blockchain, and solve the problems in related technologies that cannot satisfy the computing real-time and data requirements of IoT devices at the same time The question of security requirements.

As an optional embodiment, generating a block chain consensus mechanism according to reputation includes:

Use the target edge server to create an initial unit, where the initial unit is used to store the edge computing results and the reputation of the federation node;

using the target edge server to select a first preset number of edge blocks through a first preset method, and storing the hash values of the first preset number of edge blocks into the initial unit to obtain the first unit;

Use the target edge server to store a random number into the first unit, calculate the hash value of the first unit at this time, and store the hash value into the first unit to obtain the second unit;

using the target edge server to broadcast the second unit to other edge servers;

Use other edge servers to verify whether the second unit is legal, and if it is legal, the second unit becomes a new edge block;

Use the edge blocks generated by other edge servers to verify new edge blocks, and use other target edge servers to judge whether the number of verifications reaches the authentication threshold. If the authentication threshold is reached, the data of the second unit is successfully consensused by the cloud-native network system .

ρ_i,m为计算任务卸载策略，N为正整数。边缘服务器m为每个交易创建一个单元(Unit)，用于存储交易信息和联盟节点的信誉度，并利用私钥对此交易进行数字签名。Optionally, the edge server m performs the calculation offload task of the IoT device i, and takes the result of this calculation task as a transaction on the blockchain. At this time, the transaction arrival rate of IoT device i is λ _i , and the transaction arrival rate recorded by edge server m is

ρ _i,m is the computing task offloading strategy, and N is a positive integer. The edge server m creates a unit for each transaction, which is used to store the transaction information and the reputation of the alliance node, and uses the private key to digitally sign the transaction.

Tangle is the most representative consensus method based on the DAG blockchain. During the Tangle consensus process, each Unit needs to verify two existing edge blocks before joining the DAG blockchain ledger. (Tips), the first preset number can be two or other values, and the embodiment of the present application does not limit the specific values. When there are multiple edge blocks (more than two), the Tangle uses the Monte Carlo Markov chain (Markov Chain MonteCarlo, MCMC) edge block selection algorithm based on reputation to select two edge blocks. In the process of each edge block, the edge server generates several wandering particles, and

和

分别表示x，y和z的累计权重值。

和

分别表示发布x，y和z的节点的信誉值。z→x表示z直接引用x，α和β是非负可调参数。最先抵达边缘区块的两个粒子所停留的边缘区块即为被选中的区块。联盟节点的信誉度越高，其发布的边缘区块被选中验证的概率越高，交易验证速度越快，交易认证时延越短。当新到达的Unit选择两个边缘区块后，需检查这两个边缘区块是否冲突。如果不存在冲突，则将这两个边缘区块的哈希值存储到Unit中。Walk independently to the direction of the edge block, where,

and

denote the accumulated weight values for x, y, and z, respectively.

and

denote the reputation values of the nodes publishing x, y and z, respectively. z→x means that z directly refers to x, and α and β are non-negative adjustable parameters. The edge block where the two particles that first arrive at the edge block stay is the selected block. The higher the credibility of the alliance node, the higher the probability that the edge blocks it publishes will be selected for verification, the faster the transaction verification speed, and the shorter the transaction verification delay. When the newly arrived Unit selects two edge blocks, it needs to check whether the two edge blocks conflict. If there is no conflict, store the hash values of these two edge blocks into Unit.

In order to prevent DDoS (Distributed Denial of Service) attacks and avoid waste of computing power, the edge server needs to do a lightweight Proof of Work (PoW) before publishing the above transactions. The content and the hash value of the selected two edge blocks are packaged, and a random number Nonce is added to calculate the hash value. And store this hash value in the Unit, and then broadcast the Unit to other edge servers in the network.

When other edge servers receive the Unit, they verify whether it is legal based on the digital signature and the Nonce value. If it is legal, this Unit will be added to its local DAG ledger, and this legal Unit will become a marginal block.

The steps for other edge servers to generate edge blocks are the same as the above steps, and the edge blocks generated by other edge servers may directly or indirectly verify the above-mentioned Unit edge blocks. In the DAG blockchain, each block has a cumulative weight value (initial value is 1), and whenever the block is directly or indirectly verified, its cumulative weight value is increased by 1. When the cumulative weight value of the block reaches the authentication threshold, the transaction in the block is confirmed by the entire network as a successful transaction.

In the embodiment of this application, by designing the consensus mechanism of the DAG blockchain based on reputation, the traditional DAG consensus mechanism has the problem of prolonged transaction authentication, vulnerable to malicious node attacks, and insufficient consensus security. It effectively reduces the transaction authentication delay, improves the security of block consensus, and meets the data security requirements of IoT devices.

As an optional embodiment, obtaining the first time delay required by the consortium chain system in the calculation offloading process and the first energy consumption consumed, and obtaining the second time delay required by the blockchain consensus mechanism in the consensus process include :

Acquiring the transmission delay and first energy consumption of the task offloading information during transmission, the queuing delay of the task offloading information at the edge server, and the execution delay required by the edge server to execute the task offloading information;

Obtain the first delay according to the transmission delay, queuing delay and execution delay;

The second delay in obtaining task offloading information during the consensus process.

Optionally, in the cloud-native network, IoT devices offload computing tasks to edge servers for execution through wireless channels. During this process, transmission delay and the first energy consumption, that is, transmission energy consumption, will be generated. The uplink data transmission rate and transmission delay of the computing tasks offloaded from the IoT device d _i to the edge server S _m can be expressed as:

为其他物联网设备所产生的干扰功率，σ²为背景噪声功率。在传输阶段物联网设备d_i的能耗为

由于物联网设备的能量有限，边缘服务器有持续供电的电源，因而只考虑物联网设备的能耗，不考虑边缘服务器产生的能耗。Wherein, B is the channel bandwidth, adopts Orthogonal Frequency Division Multiplexing (OFDM), the bandwidth of each subcarrier is B ₀ =BN, and N _m is the number of terminal devices that offload computing tasks to S _m . p _i,m and H _i,m are the transmission power and channel gain from d _i to S _m respectively,

is the interference power generated by other IoT devices, and σ ² is the background noise power. The energy consumption of the IoT device d _i in the transmission phase is

Due to the limited energy of IoT devices, edge servers have continuous power supply, so only the energy consumption of IoT devices is considered, and the energy consumption generated by edge servers is not considered.

d_i在S_m上每个计算任务的平均排队时间为：The edge server has limited computing resources and cannot simultaneously perform computing offloading tasks for multiple IoT devices. When there are multiple computing offloading tasks, queuing delays will occur. This paper adopts the queuing theory, and expresses the model of the offloading tasks waiting to be executed in the edge server processing task buffer as a D/M/1 queue. According to the FCFS service rule of the IoT device task, the task arrival rate on S _m is

The average queuing time of each computing task of d _i on S _m is:

绝对值非常小，μ_m为S_m的服务速率，λ_m<μ_m。in,

The absolute value is very small, μ _m is the service rate of S _m , λ _m < μ _m .

The computing delay of each computing task d _i on S _m can be expressed as:

Among them, C _i represents the number of CPU cycles required to process 1-bit input data, and f _m represents the CPU cycle frequency of S _m .

The sum of the above transmission delay, queuing delay and calculation delay is the first delay, that is, the task unloading delay.

的泊松过程。在S_m上产生的区块的累计权重增长速率为

当此区块的累计权重值达到认证权重阈值W时，其所消耗的区块链交易认证时延即第二时延可以表示为：

In the case of an unsaturated wireless transmission environment, after the edge server completes the computing offloading task of the IoT device, it packs each computing result and related information into blocks, and performs blockchain consensus to protect the data security of computing offloading. For the blockchain network, the transaction arrival rate is related to the service rate of the edge server, and its transaction arrival rate obeys the mean value

Poisson process. The cumulative weight growth rate of blocks generated on S _m is

When the cumulative weight value of this block reaches the authentication weight threshold W, the blockchain transaction authentication delay that it consumes, that is, the second delay, can be expressed as:

In the embodiment of this application, the task unloading delay is obtained by calculating the task transmission delay, queuing delay and calculation delay, and the first energy consumption and blockchain transaction authentication delay are obtained through calculation, and the entire task is considered All time delays in the unloading process and the blockchain transaction authentication process, and with the optimization goal of minimizing the weighted sum of the task unloading delay and the blockchain transaction authentication delay, a mathematical optimization model is established to meet the needs of IoT devices. Calculate real-time requirements.

As an optional embodiment, establishing a delay minimization target model according to the first delay, the first energy consumption, and the second delay includes:

generating constraint conditions for the first time delay, the first energy consumption and the second time delay;

Based on the constraints, the delay minimization objective model is established.

Optionally, this embodiment of the present application jointly optimizes the task offloading delay of IoT devices and the blockchain consensus mechanism, and the generated constraints and established mathematical optimization model are:

Among them, ω ₁ and ω ₂ are weighting coefficients, and C1 and C2 are constraints on the time delay and energy consumption that IoT devices can tolerate, respectively. C3 is to ensure that the task arrival rate on S _m does not exceed its service rate. C4 and C5 are constraints on computing offloading decisions, ensuring that each edge server can serve multiple IoT devices, but each IoT device can only offload computing tasks to one edge server for execution. C6 is a constraint on the number of IoT devices served by each edge server.

In the embodiment of the present application, firstly, constraint conditions are generated for the time delay and energy consumption during the offloading task calculation process of the IoT device. Afterwards, with the aim of minimizing task offloading delay and blockchain transaction authentication delay, a mathematical optimization model is established based on the above constraints. Through this mathematical optimization model, an optimization scheme for computing offloading can be obtained, and a blockchain-enabled cloud The system delay of the native network is minimized, which solves the problem that related technologies cannot meet the real-time computing requirements of IoT devices.

As an optional embodiment, the calculation unloading optimization scheme obtained based on the delay minimization objective model includes:

converting the time delay minimization target model into a second preset number of partial models by using a second preset method;

optimizing the local models in parallel by training parameters of a second preset number of local models to obtain corresponding local variables;

Merge the local variables and optimize the global variables of the delay minimization objective model to obtain the optimization scheme of computing offloading.

Optionally, since the computation offloading optimization problem is not a convex problem, and with the increase of the number of users in the cloud native network, the scale and complexity of the problem will increase rapidly. This application uses the alternate direction multiplier method to transform the original complex calculation unloading optimization problem into a second preset number of unconstrained sub-problems for solution. The second preset number means multiple, and the embodiment of this application does not limit the specific value . Afterwards, computing nodes optimize subproblems in parallel by training their own local model parameters. Then, all the local variables are combined to optimize the global variables. Finally, the global solution is obtained through iteration.

In the embodiment of this application, by simultaneously optimizing the calculation offloading task and the block consensus task, and using the alternate direction multiplier method to simplify the optimization problem, the calculation offloading optimization scheme is finally obtained. The application improves the optimization efficiency, and the calculation offload optimization scheme obtained can simultaneously meet the quality of service requirements for real-time calculation and data security when IoT devices with limited massive resources frequently perform calculation offload tasks. It solves the problem that the real-time calculation and data security of the Internet of Things devices cannot be satisfied at the same time in the related technology.

As an optional embodiment, according to the edge computing results of the target edge server and the task arrival rate of the target IoT device, determining the credibility of each pair of federation nodes includes:

Using the target edge server to obtain the comprehensive reputation of the target edge server according to the credibility of the third preset number of target IoT devices to the target edge server;

Use the target edge server to obtain the reputation of the target IoT device according to the task arrival rate of the target IoT device;

According to the comprehensive reputation of the target edge server and the reputation of the target IoT device, the reputation of each pair of federation nodes is determined.

Optionally, since the edge server m can serve multiple IoT devices at the same time, its reputation should be generated by the comprehensive reputation scores of multiple IoT devices, and the comprehensive reputation is

为第三预设数量。得到综合信誉度S_i→m之后，该边缘服务器对综合信誉度进行存储。Among them, ρ _i,m is the calculation task offloading strategy, r _i→m is the score of the reputation of the edge server m according to the satisfaction of the edge server m after the IoT device i receives the calculation result of the edge server m, N is a positive integer,

for the third preset quantity. After obtaining the comprehensive reputation S _i→m , the edge server stores the comprehensive reputation.

Since the task arrival rate of the IoT device indirectly reflects the transaction arrival rate of the blockchain, the edge server m scores the reputation of the IoT device according to the task arrival rate λ _i of the IoT device S _m→i = g( λ _i ), and store S _m→i .

Taking each IoT device and its selected edge server as a pair of alliance nodes, after calculation, the reputation S _i,m of each pair of alliance nodes is obtained =αS _i→m +βS _m→i (9)

Among them, α and β are reputation weighting factors respectively, and IoT devices can adjust the weight of reputation according to service quality requirements.

In the embodiment of this application, by taking each IoT device and its selected edge server as a pair of alliance nodes, and considering the impact of their respective reputations on the performance of the blockchain, the credibility of each pair of alliance nodes is calculated to improve Block consensus security meets the quality of service requirements of IoT devices for data security.

As an optional embodiment, establishing a blockchain-based cloud native network system includes:

Establishing a blockchain-powered cloud-native network system, wherein the cloud-native network system includes a second preset number of IoT devices and a fourth preset number of edge servers;

Based on the cloud-native network system, an alliance chain system in the process of computing offload is established.

第二预设数量为N，第四预设数量为M，N和M表示多个，本申请实施例不对具体数值做限定。物联网设备d_i在执行计算密集型应用时，生成一系列相同的计算任务，物联网设备直接将上述计算任务通过无线信道传输到边缘服务器执行。假定在一个时隙内，d_i到达的任务数量为n_i＝υ/θ_i，其中，υ是一个时隙的时间长度，θ_i是任务到达的时间间隔，则λ_i＝1/θ_i为d_i的任务到达率。边缘服务器之间通过有线链路连接。区块链网络用于保证计算卸载数据的安全性。Optionally, the cloud-native network system model empowered by the blockchain is shown in Figure 2, and the system includes a cloud-native network and a blockchain network. There are N IoT devices D={d ₁ ,...,d _i ,...,d _N } and M edge servers in the cloud-native network

The second preset number is N, and the fourth preset number is M. N and M represent a plurality, and the embodiment of the present application does not limit the specific values. The IoT device d _i generates a series of identical computing tasks when executing computationally intensive applications, and the IoT device directly transmits the above computing tasks to the edge server through a wireless channel for execution. Assume that within a time slot, the number of tasks arriving at d _i is n _i =υ/θ _i , where υ is the time length of a time slot, and θ _i is the time interval between task arrivals, then λ _i =1/θ _i is the task arrival rate of d _i . The edge servers are connected through wired links. The blockchain network is used to ensure the security of computing offload data.

Build a consortium chain jointly maintained by two types of interest groups, IoT devices and edge servers. Among them, IoT devices serve as blockchain users to send task offloading information, and edge servers, as blockchain consensus nodes, perform computing offloading tasks and blockchain consensus tasks. The workflow of the consortium chain is as follows: first, the IoT device sends task offloading information to the edge server, and the edge server performs the calculation offloading task, and returns the calculation result to the IoT device after completion. Then, the edge server packs the calculation task results and related information into blocks, and completes the blockchain consensus according to the DAG-based blockchain consensus mechanism. Privacy and data security.

In the embodiment of this application, a blockchain-based cloud-native network system is established, and then an alliance chain model of IoT devices and edge servers in the process of computing offload is established according to the cloud-native network system, and the blockchain consensus mechanism is used to prevent nodes from being Malicious tampering and preventing forgery of user data information meet the security requirements of IoT devices.

According to another aspect of the embodiments of the present application, a computing offloading optimization device for implementing the above computing offloading optimization method is also provided. Fig. 3 is a structural block diagram of an optional computing offload optimization device according to an embodiment of the present application. As shown in Fig. 3, the device may include:

The establishment module 301 is used to establish a blockchain-based cloud-native network system, wherein the cloud-native network system includes an alliance chain system for information processing in the calculation offloading process, and the alliance chain system includes multiple pairs of alliance nodes, and each pair of alliance nodes Including the target IoT device and the target edge server, the target IoT device is the object that sends task offload information, and the target edge server is the object that calculates task offload information and performs blockchain consensus;

Determining module 302, for determining the credit degree of each pair of alliance nodes according to the edge calculation result of the target edge server and the task arrival rate of the target IoT device;

Generating module 303, is used for generating block chain consensus mechanism according to credibility;

The obtaining module 304 is used to obtain the first time delay and the first energy consumption required by the consortium chain system during the calculation offloading process, and obtain the second time delay required by the blockchain consensus mechanism during the consensus process;

The obtaining module 305 is configured to establish a delay minimization target model according to the first time delay, the first energy consumption, and the second time delay, and obtain a calculation unloading optimization solution based on the time delay minimization target model.

As an optional embodiment, the generation module includes:

Create a unit for creating an initial unit by using the target edge server, where the initial unit is used to store the edge computing result and the reputation of the federation node;

The selection unit is configured to use the target edge server to select a first preset number of edge blocks through a first preset method, and store the hash values of the first preset number of edge blocks into the initial unit to obtain the first unit;

The computing unit is configured to use the target edge server to store a random number into the first unit, calculate the hash value of the first unit at this time and store the hash value into the first unit to obtain the second unit;

a broadcast unit, configured to use the target edge server to broadcast the second unit to other edge servers;

The verification unit is used to use other edge servers to verify whether the second unit is legal, and if it is legal, the second unit becomes a new edge block;

The judging unit is used to use the edge blocks generated by other edge servers to verify new edge blocks, and use other edge servers to judge whether the number of verifications reaches the authentication threshold. Network system successful consensus.

As an optional embodiment, the acquisition module includes:

The first acquisition unit is configured to acquire the transmission delay of the task offloading information during transmission and the first energy consumption, the queuing delay of the task offloading information at the edge server, and the execution delay required by the edge server to execute the task offloading information;

The first obtaining unit is configured to obtain the first time delay according to the transmission time delay, queuing time delay and execution time delay;

The second acquiring unit is configured to acquire the second time delay of the task offloading information in the consensus process.

As an optional embodiment, the obtaining module includes:

A generating unit, configured to generate constraint conditions for the first time delay, the first energy consumption, and the second time delay;

The first establishing unit is configured to establish a time delay minimization target model based on constraint conditions.

A conversion unit, configured to convert the time-delay minimization target model into a second preset number of local models by a second preset method;

The second obtaining unit is configured to optimize the local models in parallel by training the parameters of a second preset number of local models to obtain corresponding local variables;

The third obtaining unit is used for merging local variables and optimizing the global variables of the time delay minimization objective model to obtain an optimization scheme of computation offloading.

As an optional embodiment, the determining module includes:

The fourth obtaining unit is used to use the target edge server to obtain the comprehensive reputation of the target edge server according to the reputation of the third preset number of target IoT devices to the target edge server;

The fifth obtaining unit is used to use the target edge server to obtain the credibility of the target IoT device according to the task arrival rate of the target IoT device;

The determining unit is configured to determine the reputation of each pair of federation nodes according to the comprehensive reputation of the target edge server and the reputation of the target IoT device.

As an optional embodiment, the establishment module includes:

The second establishment unit is used to establish a blockchain-powered cloud-native network system, wherein the cloud-native network system includes a fourth preset number of IoT devices and a fifth preset number of edge servers;

The third establishment unit is used to establish the alliance chain system in the calculation offloading process based on the cloud-native network system.

It should be noted here that the examples and application scenarios implemented by the above modules and corresponding steps are the same, but are not limited to the content disclosed in the above embodiments.

Fig. 4 is a structural block diagram of an optional electronic device according to an embodiment of the present application, as shown in Fig. 402 and memory 403 complete mutual communication through communication bus 404, wherein,

Memory 403, used to store computer programs;

When the processor 401 is used to execute the computer program stored on the memory 403, the following steps are implemented:

Establish a blockchain-based cloud-native network system, in which the cloud-native network system includes a consortium chain system for information processing during computing offloading. The consortium chain system includes multiple pairs of consortium nodes, and each pair of consortium nodes includes target IoT devices and The target edge server, the target IoT device is the object that sends task offload information, and the target edge server is the object that calculates task offload information and performs blockchain consensus;

Determine the credibility of each pair of federation nodes according to the edge computing results of the target edge server and the task arrival rate of the target IoT device;

Generate a blockchain consensus mechanism based on reputation;

Obtain the first time delay and the first energy consumption required by the alliance chain system in the calculation offloading process, and obtain the second time delay required by the blockchain consensus mechanism in the consensus process;

According to the first time delay, the first energy consumption and the second time delay, a time delay minimization target model is established, and a computation unloading optimization scheme is obtained based on the time delay minimization target model.

Optionally, in this embodiment, the aforementioned communication bus may include but not limited to a PCI (Peripheral Component Interconnect, Peripheral Component Interconnect Standard) bus, or an EISA (Extended Industry Standard Architecture, Extended Industry Standard Architecture) bus, and the like. The communication bus can be divided into an address bus, a data bus, a control bus, and the like. For ease of representation, only one thick line is used in FIG. 4 , but it does not mean that there is only one bus or one type of bus.

The communication interface is used for communication between the electronic device and other devices.

The memory may include RAM, and may also include but not limited to non-volatile memory (non-volatile memory), for example, at least one disk memory. Optionally, the memory may also be at least one storage device located away from the aforementioned processor.

As an example, as shown in FIG. 4 , the memory 403 may include, but is not limited to, the establishment module 301 , determination module 302 , generation module 303 , acquisition module 304 , and acquisition module 305 of the calculation offloading optimization apparatus. In addition, it may also include but not limited to other module units in the above computing offload optimization device, which will not be repeated in this example.

Above-mentioned processor can be general-purpose processor, can include but not limited to: CPU (Central Processing Unit, central processing unit), NP (Network Processor, network processor) etc.; Can also be DSP (DigitalSignal Processing, digital signal processor), ASIC (Application Specific Integrated Circuit, Application Specific Integrated Circuit), FPGA (Field-Programmable Gate Array, Field Programmable Gate Array) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.

Optionally, for specific examples in this embodiment, reference may be made to the examples described in the foregoing embodiments, and details are not repeated in this embodiment.

Those of ordinary skill in the art can understand that the structure shown in FIG. 4 is only schematic. FIG. 4 does not limit the structure of the electronic device. For example, the terminal device may also include, but not limited to, more or fewer components than those shown in FIG. 4 (such as a network interface, a display device, etc.), or have a configuration different from that shown in FIG. 4 .

Those skilled in the art can understand that all or part of the steps in the various methods of the above embodiments can be completed by instructing hardware related to the terminal device through a program, and the program can be stored in a computer-readable storage medium, and the storage medium can be Including but not limited to: flash disk, ROM, RAM, magnetic disk or optical disk, etc.

According to still another aspect of the embodiments of the present application, a storage medium is also provided. Optionally, in this embodiment, the foregoing storage medium may be used to store program codes for executing the computing offload optimization method.

Optionally, in this embodiment, the foregoing storage medium may be located on at least one network device among the plurality of network devices in the network shown in the foregoing embodiments.

Optionally, in this embodiment, the storage medium is configured to store program codes for performing the following steps:

Establish a blockchain-based cloud-native network system, in which the cloud-native network system includes a consortium chain system for information processing during computing offloading. The consortium chain system includes multiple pairs of consortium nodes, and each pair of consortium nodes includes target IoT devices and The target edge server, the target IoT device is the object that sends task offload information, and the target edge server is the object that calculates task offload information and performs blockchain consensus;

Determine the credibility of each pair of federation nodes according to the edge computing results of the target edge server and the task arrival rate of the target IoT device;

Generate a blockchain consensus mechanism based on reputation;

Obtain the first time delay and the first energy consumption required by the alliance chain system in the calculation offloading process, and obtain the second time delay required by the blockchain consensus mechanism in the consensus process;

According to the first time delay, the first energy consumption and the second time delay, a time delay minimization target model is established, and a computation unloading optimization scheme is obtained based on the time delay minimization target model.

Optionally, for specific examples in this embodiment, reference may be made to the examples described in the foregoing embodiments, which will not be repeated in this embodiment.

Optionally, in this embodiment, the above-mentioned storage medium may include, but not limited to, various media capable of storing program codes such as a U disk, ROM, RAM, removable hard disk, magnetic disk, or optical disk.

In the description of this specification, descriptions referring to the terms "this embodiment", "one embodiment", "some embodiments", "examples", "specific examples", or "some examples" mean that the embodiments are combined A specific feature, structure, material, or characteristic described by or example is included in at least one embodiment or example of the present disclosure. In this specification, the schematic representations of the above terms are not necessarily directed to the same embodiment or example. Furthermore, the described specific features, structures, materials or characteristics may be combined in any suitable manner in any one or more embodiments or examples. In addition, those skilled in the art can combine and combine different embodiments or examples and features of different embodiments or examples described in this specification without conflicting with each other. In the description of the present disclosure, "plurality" means at least two, such as two, three, etc., unless otherwise specifically defined.

Apparently, the above-mentioned embodiments are only examples for clear description, rather than limiting the implementation. For those of ordinary skill in the art, other changes or changes in different forms can be made on the basis of the above description. It is not necessary and impossible to exhaustively list all the implementation manners here. And the obvious changes or changes derived therefrom are still within the scope of protection of the present invention.

Claims (10)

1. A method for computing offload optimization, the method comprising:

establishing a block chain-based cloud native network system, wherein the cloud native network system comprises a alliance chain system for information processing in a computing unloading process, the alliance chain system comprises a plurality of pairs of alliance nodes, each pair of alliance nodes comprises target Internet of things equipment and a target edge server, the target Internet of things equipment is an object for sending task unloading information, and the target edge server is an object for computing the task unloading information and performing block chain consensus;

determining the credit degree of each pair of alliance nodes according to the edge calculation result of the target edge server and the task arrival rate of the target Internet of things equipment;

generating a block chain consensus mechanism according to the credibility;

acquiring a first time delay and a first consumed energy consumption required by the alliance chain system in a calculation unloading process, and acquiring a second time delay required by the block chain consensus mechanism in a consensus process;

and establishing a time delay minimization target model according to the first time delay, the first energy consumption and the second time delay, and obtaining a calculation unloading optimization scheme based on the time delay minimization target model.

2. The method of claim 1, wherein the generating a block chain consensus mechanism according to the reputation comprises:

creating an initial unit by using the target edge server, wherein the initial unit is used for storing the edge calculation result and the credibility of the federation node;

selecting a first preset number of edge blocks by using the target edge server through a first preset method, and storing the hash values of the first preset number of edge blocks into the initial unit to obtain a first unit;

storing a random number into the first unit by using the target edge server, calculating a hash value of the first unit at the moment, and storing the hash value into the first unit to obtain a second unit;

broadcasting the second cell to other edge servers using the target edge server;

verifying whether the second unit is legal by using the other edge servers, and if so, enabling the second unit to become a new edge block;

and verifying the new edge block by using the edge blocks generated by the other edge servers, judging whether the verification times reach an authentication threshold value by using the other target edge servers, and if so, successfully identifying the data of the second unit by the alliance chain system.

3. The method of claim 1, wherein the obtaining a first latency and a first energy consumption required by the federation chain system in calculating an offloading process and obtaining a second latency required by the blockchain consensus mechanism in a consensus process comprises:

acquiring transmission delay of the task unloading information in a transmission process, the first energy consumption, queuing delay of the task unloading information in a queue at an edge server and execution delay required by the edge server to execute the task unloading information;

obtaining the first time delay according to the transmission time delay, the queuing time delay and the execution time delay;

and acquiring the second time delay of the task unloading information in the consensus process.

4. The method of claim 1, wherein the establishing a latency minimization objective model according to the first latency, the first energy consumption, and the second latency comprises:

generating constraints for the first latency, the first energy consumption, and the second latency;

and establishing the time delay minimization target model based on the constraint condition.

5. The method of claim 1, wherein deriving a computational offload optimization scheme based on the latency minimization objective model comprises:

converting the time delay minimization target model into a second preset number of local models by a second preset method;

optimizing the local models in parallel by training the parameters of the second preset number of local models to obtain corresponding local variables;

and combining the local variables, and optimizing the global variable of the time delay minimization target model to obtain the calculation unloading optimization scheme.

6. The method of claim 1, wherein the determining the reputation of each pair of federation nodes according to the edge calculation result of the target edge server and the task arrival rate of the target internet of things device comprises:

obtaining a comprehensive credibility of the target edge server by using the target edge server according to the credibility of a third preset number of target Internet of things equipment to the target edge server;

obtaining the credit degree of the target Internet of things equipment by using the target edge server according to the task arrival rate of the target Internet of things equipment;

and determining the credit degree of each pair of the alliance nodes according to the comprehensive credit degree of the target edge server and the credit degree of the target Internet of things equipment.

7. The method of claim 1, wherein establishing a block chain based cloud native network system comprises:

establishing a block chain energized cloud native network system, wherein the cloud native network system comprises a second preset number of Internet of things devices and a fourth preset number of edge servers;

and establishing the alliance chain system in the computing unloading process based on the cloud native network system.

8. A computing offload optimization apparatus, comprising:

the system comprises an establishing module, a block chain-based cloud native network system and a block chain identification module, wherein the cloud native network system comprises a alliance chain system for information processing in a computing unloading process, the alliance chain system comprises a plurality of pairs of alliance nodes, each pair of alliance nodes comprises target Internet of things equipment and a target edge server, the target Internet of things equipment is an object for sending task unloading information, and the target edge server is an object for computing the task unloading information and performing block chain consensus;

the determining module is used for determining the credit degree of each pair of the alliance nodes according to the edge calculation result of the target edge server and the task arrival rate of the target Internet of things equipment;

the generating module is used for generating a block chain consensus mechanism according to the credibility;

the acquisition module is used for acquiring a first time delay and a first consumed energy which are required by the alliance chain system in the process of calculating and unloading, and acquiring a second time delay required by the block chain consensus mechanism in the consensus process;

and the obtaining module is used for establishing a time delay minimization target model according to the first time delay, the first energy consumption and the second time delay and obtaining a calculation unloading optimization scheme based on the time delay minimization target model.

9. An electronic device comprising a processor, a communication interface, a memory and a communication bus, wherein said processor, said communication interface and said memory communicate with each other via said communication bus,

the memory for storing a computer program;

the processor for performing the method steps of any one of claims 1 to 7 by running the computer program stored on the memory.

10. A computer-readable storage medium, in which a computer program is stored which, when being executed by a processor, carries out the method steps of any one of claims 1 to 7.

Images